Comparison between the deconvolution and maximum slope 64-MDCT perfusion analysis of the esophageal cancer: Is conversion possible?

Abstract

Purpose: To estimate if CT perfusion parameter values of the esophageal cancer, which were obtained with the deconvolution-based software and maximum slope algorithm are in agreement, or at least interchangeable. Methods: 278 esophageal tumor ROIs, derived from 35 CT perfusion studies that were performed with a 64-MDCT, were analyzed. "Slice-by-slice" and average "whole-covered-tumor-volume" analysis was performed. Tumor blood flow and blood volume were manually calculated from the arterial tumortime- density graphs, according to the maximum slope methodology (BFms and BV ms), and compared with the corresponding perfusion values, which were automatically computed by commercial deconvolutionbased software (BF deconvolution and BVdeconvolution), for the same tumor ROIs. Statistical analysis was performed using Wilcoxon matched-pairs test, paired-samples t-test, Spearman and Pearson correlation coefficients, and Bland-Altman agreement plots. Results: BFdeconvolution (median: 74.75 ml/min/100 g, range, 18.00-230.5) significantly exceeded the BF ms(25.39 ml/min/100 g, range, 7.13-96.41) (Z = -14.390, p < 0.001), while BVdeconvolution (median: 5.70 ml/100 g, range: 2.10-15.90) descended the BVms(9.37 ml/100 g, range: 3.44-19.40) (Z = -13.868, p < 0.001). Both pairs of perfusion measurements significantly correlated with each other: BFdeconvolution, versus BFms (rS = 0.585, p < 0.001), and BVdeconvolution, versus BV ms (rS = 0.602, p < 0.001). Geometric mean BF deconvolution/BFms ratio was 2.8 (range, 1.1-6.8), while geometric mean BVdeconvolution/BVms ratio was 0.6 (range, 0.3-1.1), within 95% limits of agreement. Conclusions: Significantly different CT perfusion values of the esophageal cancer blood flow and blood volume were obtained by deconvolution-based and maximum slope-based algorithms, although they correlated significantly with each other. Two perfusion-measuring algorithms are not interchangeable because too wide ranges of the conversion factors were found. © 2013 Elsevier Ireland Ltd.